Tidy data §

Attributes §

• 3rd normal form
  • each variable -> one column
  • each observation -> one row
  • each type of observational unit -> one table
    • observational unit = e.g. a questionnaire

Common Problems §

• Column = value instead of variable name (e.g. col = “pregnant” and “not pregnant”)
  • -> long format (pivot_longer)
• Column has more than one variable (e.g. col = “num-male-inf-students”)
  • -> separate (separate)
• Variable is in both rows and columns (in a field)
  • -> wide format (pivot_wider)
• Multiple observation units in one table (e.g. two different questionnaires, repeating values)
• (One observation split over multiple tables)

Todo §

1. Small cleanup §

• post_study_questionnaires$knows_about_featured_snippets fac

• post_study_questionnaires$featured_snippet_usage fac

• pre_task_questionnaires$pre_answer fac

• post_task_questionnaires$post_answer fac

• snippetscochrane_answer & anno_final fac

2. Joins & new tables §

• create table topic_results
  • $use{r}_{i}d,$topic, $time (from pre/post_task_questionnaires)
• move demographics$stud{y}_{e}n{d}_{t}ime->$users and calc time

3. Further cleanup §

• create web_page_results$doc_time (start-end)